MapReduce Batch Jobs Approch for Data Analysis

MR1 (finding the Average Book Rating)

 input: BX-Book-Ratings.csv
 Mapper - gets values of ISBN and BookRating and send it in reducer phase, key= ISBN and value= Ratings
 Reducer - get key & iterate it's value = get values and find average.

MR2 (getting most popular ISBN locations) (where are books most popular)

input: BX-Book-Ratings.csv and BX-Users.csv

 Mapper1- Maps values from both the csv file. (by matching the pattern ".*\\d+.*")
 BX-Book-Ratings.csv: get userid, ISBN (key,value)
 BX-User.csv: get userid, location (key, value)

(one userid can reference multipe ISBNs, but single location)

 Reducer-1 - create new key, from ISBN & userlocation, and its value part will contain user ids
 Mapper-2 - get the key and value, from reducer-1,
 Reducer-2 - get new key (ISBN+userlocation) and iterate its values.

each iterate will increment the count by 1, (for counting nuber of userids it has) (final output will contain ISBN+userlocation & usercounts)

this will indicate which ISBN is famous at which location.

MR3 (how many books each publisher has published every year)

 Mapper - key: yearofpublish + publishername value: 1
 Reducer - get key and iterate through values and sum the count.

MR4 (ISBN vs Age group Analysis) (Which book(ISBN) is most sold in every age-group?)

input: BX-Book-Ratings and BX-Users

 Mapper-1 - key: userID value: ISBN
             key: userID value: age

(one user ID can have multiple ISBN but single Age Value)

Reducer-1 - key: ISBN+agegroup value: 1 AGEGROUP:

 if(age <= 19){
ageGroup = "Teenager";
}
else if(age19 && age <=34){
ageGroup = "Millenial";
}
else if(age34 && age <=50){
ageGroup = "GenX";
}
else if(age50 && age <=69){
ageGroup = "Boomer";
}
else{
ageGroup = "Silent";
}

 Mapper-2 - key: ISBN+agegroup, value:1
 Reducer-2 - get the key, iterate through values and sum count.

MR5 (How many books have each of these Authors published every year?)

same as MR3 but here we are considering authors not publishers.

input: BX-Books.csv

 
  Mapper - key: yearofpublish + author, value: 1
 Reducer - get the key and iterate through value and sum count.

MR6 (How many books have been released per year?)

 Mapper - key: year value: 1
 Reducer - get the keys, iterate through values and do sum count.